opentargets/curation: Patch release for 26.09 platform release

Le résumé fourni par la source

Patch release for the 26.09 platform release. Clears a stray value in the ancestry column on two rows of the Autism Sequencing Consortium cohort (PMID 31981491): KCNMA1 (ENSG00000156113) and CORO1A (ENSG00000102879) both carried ancestry: "5.00E-03". The value did not match either row's own p-value (pValueMantissa: 2.34, pValueExponent: -3), so it was not a shifted copy of that field — most likely a value pasted into the wrong column during curation. Both rows now use the blank convention followed by the rest of the cohort. Impact on pipeline output: none beyond the column itself. The stray value never resolved to an ancestryId, and ancestryId is what gene_burden deduplication keys on — so row counts, evidence scores and directionOnTrait are all unchanged. These are the only two rows in the dataset whose ancestry parses as a number. Full changelog: https://github.com/opentargets/curation/compare/26.08.2...26.08.3

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.